Explore UCD

UCD Home >

Using machine learning to identify “greenwashers”

Wednesday, 19 July, 2023

Andreas Hoepner (pictured) is Full Professor of Operational Risk, Banking and Finance at the Michael Smurfit Graduate Business School and the Lochlann Quinn School of Business at University College Dublin. Here he speaks about sustainable finance and the role of machine learning at(opens in a new window)greenwatch.ai, a tool for the financial services sector to assess and monitor the authenticity of green claims made by companies.

The old adage that money makes the world go round might depend these days on where it is invested. More specifically, future-proofing the planet against the threat of climate change very much relies on what stops being financed.

“The Paris agreement can be boiled down to one thing: de-browning. If we continue to put fresh billions into coal, oil and gas for further exploration - not just continuing what they're doing but further exploration, thereby worsening the situation - we have no chance at all of achieving the Paris Agreement,” says Professor Andreas Hoepner, of UCD School of Business. “We need to stop further exploration and gradually downsize current operations.”

Despite this stark reality, there is still, incredibly, a tap of money pouring into these industries.

“There is a lot of attention on, first of all, making people realise that that tap is still open and considerably open. TheFinancial Times, for instance, recently covered one piece of our research pointing out a couple of asset managers which are giving(opens in a new window)400% more fresh cash to coal.”

The Paris Agreement is an international, binding treaty whose goal is to limit global warming increases within 1.5 degrees Celsius above preindustrial levels. This requires decreasing carbon pollution by 45% from 2010 levels by 2030 and reaching net-zero carbon emissions by 2050.

Sustainable finance plays arguably the most important role in decarbonising the world, by moving money away from polluters and into greener initiatives.

“Sustainable finance’s objective is that the financing of companies and projects is preferably done in a way that leads to climate change mitigation rather than climate change worsening.”

He was encouraged by some tangible progress that arose from last month’s COP27 meeting at Sharm El Sheikh.

“The UN put out a really good(opens in a new window)reportfrom a high level expert group on what companies need to do about climate change and what they should not do. Thereby they effectively provided a recipe to identify which companies are greenwashers.”

Greenwashers are companies that “pretend to be green”, presenting themselves as climate-friendly and potentially carbon neutral, while actually not decreasing their CO2 emissions at all or at a sufficient rate.

“If you go to greenwatch.ai you'll find names,” says Prof Hoepner of the watchdog website he runs with a team from UCD and which uses artificial intelligence to speed up the process of analysing how companies’ various activities align, or not, with green goals. 

“In the GreenWatch project we're using machine learning to suggest to what extent statements represent greenwashing. Machine learning can suggest individual actions of, say, misbehaviour of companies. That would be the way that we're using it,” he explains.

“ Machine learning is predominantly machine imitation. I mean, the machine itself doesn't actually necessarily think at this stage but it can imitate human behaviour in a very highly scalable process and thereby it can be extremely cost efficient. And so, in that sense, machine learning can help around the world at scaling a whole range of activities. ”

A headline from COP27 estimated that $1 trillion of climate finance is needed per year by 2025 to cut greenhouse gases, boost climate resilience, restore nature and pay for the damage caused by extreme weather events in developing countries outside of China alone. 

“It's a conceptual sum of money,” argues Prof Hoepner, calling such headlines “clickbait”. He says quitting investment in the fossil fuel industry “is entirely free because investors can just move their money elsewhere. And in the second stage, it is this idea of greening the economy at the same time, so that the de-browning doesn't lead to any reductions in quality of life.” While the second stage is “potentially very expensive” it will hopefully be off-set by myriad innovations and opportunities to more safely support a high quality of life.

He points to companies that are already simultaneously successful and climate friendly, such as electric car manufacturer Tesla, which “tends to be worth more than the next five car manufacturers put together”. 

Apart from switching to electric or plug-in hybrid cars - not self-charging hybrids which are “a marketing trick” - he recommends individuals make a point of asking their pensions and insurance companies not to invest in companies harming the planet.

Investment companies are “not technically obliged, but they’re commercially encouraged” to invest in green companies. 

Meanwhile, every year fossil fuel companies need fresh billions to continue digging and drilling, activities that worsen climate change. 

“Most of the financing that goes to coal, oil and gas is our money as societies; our pensions or insurances. Closing that financing tap just means that everyone whose money that is realises, ‘Actually, I don't want that’. So it's a bit of an educational task to explain to people that the reason we have climate change is because our money as society - either our pensions or the money we give to an insurance trust - is moved still in the billions to pay for more oil or more gas from the ground and even processes like coal mining. And if that changes, then we are on a pretty good path.”

The European Union has adopted what it calls the Paris-aligned benchmarks, which are guidelines for greener investing. 

“Not every pension fund is forced to do it. Many of the public pension funds are aiming to do it. But everyone can talk to their pension fund and say, ‘Are you investing in line with the Paris-aligned benchmark?’ It's a very easy question to ask to make sure that we are not supporting that tap of money. If we close that financing tap then we can actually come pretty close to the Paris Agreement. Maybe we can achieve it.” 

Listen to the (opens in a new window)podcast